Optimal content-dependent dynamic brightness scaling for OLED displays

Brightness scaling is the most common way to reduce power consumption in OLED displays. Such “dimming” is generally static, i.e. it is applied either manually by the user, or automatically by the system in correspondence of predefined battery state-of-charge conditions. This is obviously sub-optimal, because it makes brightness adaptation (i) too coarse-grain in time, and (ii) agnostic of the image being displayed. In this work, we overcome the two above limitations by proposing a novel brightness scaling approach based on the online calculation of a content-dependent optimal dimming factor. While keeping the simplicity of traditional scaling, this solution enables a true dynamic scaling applicable on a frame-by-frame basis. Results show that the proposed strategy is able to obtain an average power saving greater than 30% for different reference image datasets, while maintaining the Mean Structural Similarity Index (MSSIM) between the original and transformed images at about 97%. We also perform an analysis of the costs for both hardware and software implementations of the transformation.